Functional identification of unknown proteins discovered in genome projects remains a major challenge for contemporary biology. This Program Project is focused on developing an integrated strategy for (nontrivial) functional assignment of unknown enzymes by predicting the substrate specificities of members of the functionally diverse enolase, amidohydrolase (AH), and D-ribulose 1,5-bisphosphate carboxylase/oxygenase (RuBisCO) superfamilies that share the ubiquitous (p/a)8-fold. In the past project period, the Program Project brought together expertise in computational enzymology (bioinformatics, homology modeling, and molecular docking), structural enzymology (high resolution x-ray structural analysis), and functional enzymology (protein purification, measurement of function, and determination of mechanism). We demonstrated that accurate computational prediction of substrate specificities of uncharacterized enzymes is possible using either an experimentally determined structure or a homology model, thereby facilitating experimental verification of function. In this competing renewal application, a new focus is on unknown members of the enolase, AH, and RuBisCO superfamilies that participate in novel metabolic pathways as deduced by operon context. This new focus adds to our previous "one enzyme-one function" approach and is based on the expectation that enzymes that occur in the same metabolic pathway will share conserved elements of substrate specificity, facilitating functional assignment of not only the unknown superfamily "targets" but, also, the entire metabolic pathway. The impact of this approach is considerable because it will identify new enzymes, new metabolites, new pathways, and, therefore, new biology. The integrated strategy developed in this Project will be applicable to deciphering the ligand specificity of any uncharacterized enzyme. The goals of this Program Project extend the contribution of the Protein Structure Initiative funded by NIGMS that seeks to obtain structures for proteins of unknown function that will allow reliable homology modeling.